United States

This paper presents the NIT-Agartala-NLPTeam&#39;s submission to SemEval-2026 Task 9
on polarization detection in textual data. The
task comprises two subtasks: (i) binary classification to distinguish polarized from nonpolarized content, and (ii) multi-label classification to identify the specific type(s) of polarization. We propose a weighted soft-voting
ensemble framework that integrates multiple
fine-tuned large language models (LLMs). The
probabilistic outputs of the individual models are combined using weighted averaging
to effectively leverage their complementary
strengths and enhance overall performance.
Our system achieved a test macro F1-score of
78.6 (26th out of 44 teams) in Subtask 1 and
46.0 (18th out of 29 teams) in Subtask 2.

ACL 2026

NIT-Agartala-NLP-Team at SemEval-2026 Task 9: A Weighted Soft-Voting Ensemble Framework of Fine-Tuned LLMs for Binary and Multi-Label Polarization Detection

threshold optimization

fine tuning

ensemble

It is my great pleasure and honor to welcome you to the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), held in San Diego, California, USA, July 2–7, 2026. ACL 2026 continues our fields tradition of excellence in scholarship, innovation, and inclusivity, and I am deeply grateful to the many volunteers who have worked tirelessly to bring this event to life.

It is exciting to see the many developments in the field reflected at our conference, such as the advancement of large language models (born out of our community) into many application directions. This is not without challenges, partly due to unsolved problems, and partly due to misuse of the technology. All this is reflected well at ACL 2026. The conference also reflects that the field is not only about advancing technology, but also the science of fundamentally understanding language and how it is use by many different speaker communities. <br>
[Full Message from General Chair Philipp Koehn](https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing) <br><br><br>
<html><button style="display: inline-flex; align-items: center; justify-content: center; white-space: nowrap; border-radius: 9999px; font-weight: bold; background: #7c3aed; color: white; font-family: 'Space Grotesk', sans-serif; height: 40px; font-size: 16px; padding: 0 20px; border: none; cursor: pointer" onclick="window.open('https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing')">Conference Handbook</button></html>
<br><br><br><br>

To access this event page, you need to log in with the **email address you registered with**. <br>Access credentials will be sent to your email from Underline -  subject line "Welcome to ACL 2026". Please be sure to check your spam email folder if you do not see an email confirmation right away.

To access this event page, you are required to register.
Please complete your registration to continue.

We recommend reading [**the registration information**](https://2026.aclweb.org/registration/) first.

Registration Required

The 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026) will take place in San Diego, California, from July 2 - 7, 2026.

This paper presents the NIT-Agartala-NLPTeam's submission to SemEval-2026 Task 9
on polarization detection in textual data. The
task comprises two subtasks: (i) binary classification to distinguish polarized from nonpolarized content, and (ii) multi-label classification to identify the specific type(s) of polarization. We propose a weighted soft-voting
ensemble framework that integrates multiple
fine-tuned large language models (LLMs). The
probabilistic outputs of the individual models are combined using weighted averaging
to effectively leverage their complementary
strengths and enhance overall performance.
Our system achieved a test macro F1-score of
78.6 (26th out of 44 teams) in Subtask 1 and
46.0 (18th out of 29 teams) in Subtask 2.

workshop paper

We present our coarse-to-fine Valence-Arousal (VA) ensemble system for subtask 1 of task 3 (DimABSA) which covers aspect-level VA prediction. We use a pair of trained Qwen 3 8B LoRA-tuned LLMs to predict coarse bins between 1 and 8, providing ordinal VA guidance signals along with distributional features. We then train an instruction-style, multilingual E5 encoder model with a multitask head using these LLM-derived guidance features to produce continuous VA predictions. At inference time, the same guidance signals are generated for the test set by the trained LLMs and fed into the trained encoder. This approach leverages the LLM as a high-level prior while relying on the encoder for precise calibration across languages and domains. Our system achieves an RMSE_VA of 1.20 across six languages and five domains. We compare the joint VA model to separated valence and arousal models trained on coarsened ground truth data, showing that it outperforms them, particularly on arousal correlations.

RPI Team at SemEval-2026 Task 3: An LLM-Encoder Ensemble for Coarse-to-Fine Valence-Arousal Sentiment Prediction

We present a Span-Consistency Network (SCN) for conspiracy marker extraction in English social media text. The task requires identifying character-level spans for five marker types (Actor, Action, Effect, Evidence, and Victim) under overlap-based Macro F1 evaluation. Standard token-level classifiers often produce fragmented spans, ignore inter-marker dependencies, and struggle with severe class imbalance.

Our approach addresses these challenges through three components. First, a Span Consistency Layer (SCL) propagates span-level confidence signals to encourage coherent boundary formation. Second, Cross-Marker Attention (CMA) models co-occurrence patterns between marker types via a learned correlation matrix. Third, we introduce Span Count Regularization (SCR), a total-variation-based constraint that aligns soft token probabilities with the expected number of discrete spans, mitigating prediction collapse under threshold decoding.

Built on DeBERTa-v3-large and trained with a recall-biased Tversky loss, our system is ensembled across five stratified folds. It achieved a Macro F1 of 0.24 on the official test set, placing second among participating teams. Ablation studies show that SCR plays a critical role in maintaining span structure, particularly for low-frequency and long-span markers.

CredenceAI at SemEval-2026 Task 10: A Span-Consistency Network with Cross-Marker Attention for Conspiracy Marker Extraction

This paper presents our findings for SemEval-2026 Task 9. We submit to all three subtasks using an LLM-as-an-annotator strategy, simulating the data annotation process with large language models. We created 30 LLM annotators using persona injection (also known as sociodemographic prompting) and experimented with various annotation aggregation methods, including Dawid-Skene and MACE. To further increase the variability in annotator responses, we used the hatefulness detection task as proxy for identifying polarisation. Our findings indicate that this reframing of the problem is effective for the binary classification of polarisation, but is less effective for finer-grained polarisation detection. For subtasks 2 and 3, majority voting yielded the best overall performance. While our unsupervised approach does not rank as highly as supervised ones, this work provides insight into the utility of persona-based prompting and the issue of LLM annotators exhibiting high intra-model agreement.

ShefFriday at SemEval-2026 Task 9: LLM-Based Annotation Methods for Detecting Multilingual, Multicultural and Multievent Online Polarisation

Conspiracy believers use significantly more psycholinguistic markers per post than nonbelievers (Cohen's d = 0.56, p < 10⁻⁸⁰), a pattern we term narrative density, suggesting that belief manifests as structurally denser conspiratorial frames distributed across the full text rather than concentrated in specific lexical cues.

We present Truth Gradient's system for SemEval-2026 Task 10 Subtask 2 (Samory et al., 2026): a DeBERTaV3-large model with mean pooling and a 5-seed probability-averaging ensemble achieving macro F1 = 0.829 on the 77-sample development set and 0.75 on the official test set. The 5-fold CV estimate (0.734 ± 0.007) proves the more reliable predictor of test performance, and we recommend it as standard practice for low-resource shared tasks.

Two convergent tests support the narrative density account: masking annotated marker spans drops F1 by 5.3 pp, and direct marker-count fusion recovers +0.9 pp, though we note these are not conclusive given the small dev set. Cross-validated ablation identifies encoder fine-tuning as the dominant design factor (−7.2 pts), and layer-wise probing confirms belief information peaks at mid-stack layers (layer 16/24).

Truth Gradient at SemEval-2026 Task 10:Conspiracy Belief Detection via Narrative Density and Mean Pooling

Conspiracy theories widely spread on social media and can harm society 
by increasing mistrust, vaccine hesitancy, and political radicalization. 
However, most automated detection systems have traditionally relied on 
topic-specific classifiers, which often struggle to generalize across 
domains and provide little explanation for why a text is considered 
conspiratorial. To address these limitations, this paper explores various 
LLMs on the SemEval-2026 Task 10: psycholinguistic conspiracy marker 
extraction and binary conspiracy detection from Reddit submission 
statements. Specifically, we adopt a training-free few-shot prompting 
approach using different instruction-tuned large language models under a 
variety of few-shot settings (k in {0,1,5,10,15, 20}). Within this 
framework, the proposed prompting strategy incorporates psychology-informed 
instructions to guide the models in identifying conspiracy-related signals. 
As a result, the presented system achieves an F1 score of 0.21 for marker 
extraction and 0.81 for conspiracy detection, ranking 16th out of 30 teams 
in Subtask~1 and 36th out of 52 in Subtask~2 without any task-specific 
fine-tuning. These results suggest that psycholinguistically grounded 
prompting can support interpretable conspiracy analysis; however, challenges 
remain in identifying implicit markers.

CUET_320 at SemEval-2026 Task 10: Few-Shot Large Language Models for Psycholinguistic Marker Extraction and Conspiracy Detection

Despite the need to address the proliferation of conspiracy theories in online discussions, there is a lack of benchmarks for effectively detecting conspiracy-related content in everyday conversational settings. We introduce a novel dataset of comments from Reddit, ranging from politics to TV series, as well as two synergetic tasks: (1) extracting five psycholinguistic markers, grounded in evolutionary psychology, and (2) detecting conspiracy content. The data enable multi-task approaches, allowing testing of whether marker extraction improves detection performance.

SemEval-2026 Task 10: PsyCoMark -- Psycholinguistic Conspiracy Marker Extraction and Detection

We present our shared task on evaluating the adaptability of LLMs and NLP systems across multiple languages and cultures. The task data consist of an extended version of our manually constructed BLEnD benchmark (Myung et al., 2024), covering more than 30 language--culture pairs, predominantly representing low-resource languages spoken across multiple continents. As the task is designed strictly for evaluation, participants were not permitted to use the data for training, fine-tuning, few-shot learning, or any other form of model modification.

Our task includes two tracks: (a) Short-Answer Questions (SAQ) and (b) Multiple-Choice Questions (MCQ). Participants were required to predict labels and were allowed to submit any NLP system and adopt diverse modelling strategies, provided that the benchmark was used solely for evaluation. The task attracted more than 140 registered participants, and we received final submissions from 62 teams, along with 19 system description papers.

We report the results and present an analysis of the best-performing systems and the most commonly adopted approaches. Furthermore, we discuss shared insights into open questions and challenges related to evaluation, misalignment, and methodological perspectives on model behaviour in low-resource languages and for under-represented cultures. Our data and resources are available at https://github.com/BLEnD-SemEval2026/SemEval-2026-Task-7.

SemEval-2026 Task 7: Everyday Knowledge Across Diverse Languages and Cultures

Medical query rewriting transforms verbose
consumer health questions into concise clin-
ical queries, a critical step in health infor-
mation retrieval. Large language models
(LLMs) perform well on this task by stan-
dard metrics, yet high ROUGE or BERTScore
does not guarantee preservation of clinical
content. To address this issue, we intro-
duce MedFaith-F1, a category-level faithful-
ness metric over four clinically salient cate-
gories: clinical problems, medications, proce-
dures, and follow-up intent. We further propose
a hybrid Evidence and Knowledge-Grounded
Retrieval-Augmented Generation (EKG-RAG),
an evidence and knowledge-grounded frame-
work combining hybrid retrieval over PubMed
and MedlinePlus resources with UMLS (Uni-
fied Medical Language System)-aligned ontol-
ogy grounding. Evaluating LLMs LLaMA-
3 and Qwen2.5 across zero-shot, few-shot,
and QLoRA settings on MeQSum and med-
ical question-pair (MQP) datasets revealed
that base models exhibit category-level faith-
fulness failure rates (CHR) exceeding 40%,
invisible to standard metrics, while EKG-
RAG with QLoRA reduces CHR to 26.75%,
achieving MedFaith-F1 of 0.73. Our find-
ings call for faithfulness-aware evaluation
in clinical query rewriting, and MedFaith-
F1 provides a reproducible step in that di-
rection.

Beyond Lexical Similarity: Evaluating Faithfulness in LLM-Based Medical Question Reformulation

This paper presents a hybrid solution for a binary classification of medical PubMed articles created for identifying reports that associate clinical metadata with SARS-CoV-2 genomic sequences. The system is designed to catch the subtle distinction between reports of sequence-associated patient metadata and sentences where such metadata is either unrelated, irrelevant, or linked to previous studies. The biggest challenge is the fact that the medical dataset is highly imbalanced, consisting of only 13.3 % of medical reports labeled positive. Our system proposes a hybrid solution combining four approaches that include dual-evidence tagging, negation-aware suppression, semantic frame extraction, and adversarial training. All these approaches were tested on multiple models: BiomedBERT-baseabstract, BioLinkBERT-large, PubMedBERTbase-fulltext, followed by a best-subset ensemble search to obtain an F1 score of 0.786, setting a new benchmark and positioning the solution in 1st place in the competition.

MetaMiners at SMM4H-HeaRD 2026: A Semantic-StructuralKnowledge-Enriched Ensemble for SARS-CoV-2 Metadata Identification

This paper describes the BIT.UA sys-
tem for the MultiClinNER shared task at
#SMM4H–HeaRD 2026, targeting multi-
lingual clinical named entity recognition
across seven languages for three entity types
(Disease, Procedure, Symptom). We extend
the Multi-Head CRF architecture, originally
developed for multi-class NER on Spanish
clinical text, to the multilingual setting. To
enable joint multi-entity training despite per-
entity text variations in the dataset, we develop
an adaptive text consolidation pipeline that
preserves over 94% of annotations. Our central
finding is that a single xlm-roberta-large model,
trained jointly on all seven languages and
three entity types, achieves competition rank 2
for five of seven languages, outperforming
dedicated monolingual models by up to +6.94
F1 points, while requiring only a single set
of weights. Ensembling multiple seeds of
this model achieves rank 1 for those five
languages, and combining it with monolingual
models yields rank 1 for the remaining
two.

Content not yet available

Downloads

Next from ACL 2026

RPI Team at SemEval-2026 Task 3: An LLM-Encoder Ensemble for Coarse-to-Fine Valence-Arousal Sentiment Prediction

Similar lecture

The Counterfactuals at SemEval-2026 Task 9: Can Counterfactually-Inspired Preprocessing help Detect Polarization?

Content not yet available

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from ACL 2026

RPI Team at SemEval-2026 Task 3: An LLM-Encoder Ensemble for Coarse-to-Fine Valence-Arousal Sentiment Prediction

Similar lecture

The Counterfactuals at SemEval-2026 Task 9: Can Counterfactually-Inspired Preprocessing help Detect Polarization?

Downloads